The Bank of England has issued a warning in its latest assessment, highlighting that artificial intelligence is increasingly posing risks to financial stability. The central bank notes sustained market funding is being directed towards AI's development prospects, while the technology also significantly raises the probability of cyber attacks on the banking sector.
The semi-annual Financial Stability Report points out that traditional risks within the UK financial system, such as elevated equity market valuations, substantial public debt, and high-risk private credit to businesses, have not dissipated. Compared to the previous assessment, this round introduces several new risk variables. These include increased borrowing by institutions like hedge funds to boost equity holdings, AI companies relying on large-scale borrowing to sustain expansion, and the rapidly growing potential disruptive power of artificial intelligence.
The Bank of England emphasized that for investments in the AI sector to yield returns, several conditions must be met. These include widespread profitability across the industry, the deployment of necessary supporting infrastructure, and the continued availability of financing channels.
The central bank warned that a market reassessment, leading to a downgrade of optimistic expectations for the AI industry, could easily trigger a sharp decline in stock prices. Currently, market positioning is highly concentrated. The combination of significant trend-following positions and high leverage could amplify market volatility during a downturn. The future profitability prospects of AI firms are also directly linked to the safety of their debt. Insufficient transparency regarding borrowing by these related companies could amplify risk transmission during a crisis period.
Regulatory bodies worldwide are now focusing on risks stemming from artificial intelligence. Advanced large language models and autonomous intelligent systems are substantially altering the security environment of financial operations. A recent statement by a Bank of England deputy governor indicated a need for specific regulatory rules to address risks posed by autonomous agents. The existing regulatory framework is not suited for highly autonomous intelligent systems, and relying solely on human oversight throughout the process is impractical.
The Bank of England also noted that at this stage, it remains unclear whether artificial intelligence will empower cyber defenses or facilitate cyber attacks to a greater extent. The financial industry's need to frequently update software systems to adapt to intelligent technology introduces new operational risks, such as service disruptions, inherent in the software iteration process itself.
Comments